game recording for analysis. Good or bad idea?

Japanese Reach Mahjong Rules. Strategy, news, sets - anything!

Moderator: Shirluban

Senechal
Senior Reacher
Senior Reacher
Posts: 175
Joined: Wed Mar 04, 2009 5:42 am
Location: The frozen part of HELL!

Re:game recording for analysis. Good or bad idea?

Post by Senechal » Tue May 19, 2009 11:19 pm

Tha_Reaper wrote: on the tenyao comment: i find myself finishing with tenyao hands way too often. with the vid that i\'m going to make, i hope that some people can help me in avoiding such hands, and can teach me some ways to go for the higher-point hands that i didnt think of before.
Tanyao in and of itself isn\'t a problem, it\'s how novice players tend to use it. Making 1000 points to block potential called mangan++ hands is always a good idea but if all you do make is 1000 or 2000 points a hand because you open too fast or can\'t attempt to throw in even an open sanshoku (triple straight: 345p, 345s, 345m) in the mix it\'s because in the long run, these plays lose out to people who play with closed hands.
Club Riichi de Montréal (Canada) http://riichi.ca/ (If you're from elsewhere, keep in touch with us too!)

HotelFSR
Expert Reacher
Expert Reacher
Posts: 328
Joined: Sat Jan 17, 2009 11:39 am
Location: Federated States of Micronesia

Re:game recording for analysis. Good or bad idea?

Post by HotelFSR » Wed May 20, 2009 2:12 am

Here are the videos as promised. I\'ll repost them in their own thread as well.

Part 1:

[video]http://www.youtube.com/watch?v=wBhqITfp6so[/video]



Part 2:

[video]http://www.youtube.com/watch?v=72urknssNQM[/video]


EDIT:

1.1 1:30. Normally I would never Kan, but it was late in the hand and I thought I could push a little and maybe even win since I had the possibility of three concealed triples. More importantly, though, the dead wall draw is replenished from the live wall, so I was actually robbing him of another (final) self-draw in the endgame.

Given the number of turns left, I was effectively reducing his self-draw chances by about 25%, which is actually more than his chances of hitting another ura-dora (about 15%) so the tradeoff is worth it. Gotta play the numbers, although my plan changed when the chi presented itself for a quick win and the possibility of double ron against the dealer, who is unlikely to fold.

1.2 7:00. I did not get rid of the red five early because I prefer to keep dangerous tiles with a weak hand. I can use them in the event my hand improves, but I like to keep the opponents guessing where the dora tiles are. It keeps the element of uncertainty and they might wait for a red five in vain. Either way, it didn\'t look from their early piles like they needed it. The bluff value of an early strong discard will have no effect on most players, and that was a normal level Tenhou table.

1.2 8:36. I was keeping the east and sevens as safe tiles against the other players, who seemed to still be pushing. I always keep an eye on the other players and I feared they might be ready or very close. Sometimes you want your safest tiles for the very end when everyone will be tenpai.

Bunta
Fresh Reacher
Fresh Reacher
Posts: 37
Joined: Sat Jan 17, 2009 10:16 am

Re:game recording for analysis. Good or bad idea?

Post by Bunta » Wed May 20, 2009 10:42 am

First i have to say that i think its a great idea to do this video analysis. You can definitly learn a lot if there are good videos.
Nice job hotel, i guess it was pretty hard to do a live comment on a game, but i have to say somethings about your videos.

in 1.1 at 1.30 min the south player reached. You read his wait with 3-6 so you cant get rid of the 3man and your folding your hand, but why did you make a kan? it gives him 2 more dora in case of winning and if he tsumo you eventually have to pay a lot more if he hits one of them.

Why didnt you throw the red five in 1.2 at 7 min immediatly? cuz its is definitly safe and the later the game the higher the chances of the other players to use this red five (although their discards dont look like that).

at 1.2 8.36 min you throwing a 8man which is not absolutly safe although you have another 7sou and the east(not shady, cuz its already been thrown after the reach and safe against south and west cuz its in their discards) as absolutly safe tiles.

But all in all well done Hotel!

HotelFSR
Expert Reacher
Expert Reacher
Posts: 328
Joined: Sat Jan 17, 2009 11:39 am
Location: Federated States of Micronesia

Re:game recording for analysis. Good or bad idea?

Post by HotelFSR » Wed May 20, 2009 11:13 am

Thanks for the feedback.

1.1 1:30. Normally I would never Kan, but it was late in the hand and I thought I could push a little and maybe even win since I had the possibility of three concealed triples. More importantly, though, the dead wall draw is replenished from the live wall, so I was actually robbing him of another (final) self-draw in the endgame.

Given the number of tiles left, I was effectively reducing his self-draw chances by about 25%, which is actually more than his chances of hitting another ura-dora (about 15%) so the tradeoff is worth it. Gotta play the numbers, although my plan changed when the chi presented itself for a quick win and the possibility of double ron against the dealer, who is unlikely to fold.

1.2 7:00. I prefer to keep dangerous tiles with a weak hand. I can use them in the event my hand improves, but I like to keep the opponents guessing where the dora tiles are. It keeps the element of uncertainty and they might wait for a red five in vain. Either way, it didn\'t look from their early piles like they needed it. The bluff value of an early strong discard will have no effect on most players, and that was a normal level Tenhou table.

1.2 8:36. I was keeping the east and sevens as safe tiles against the other players, who seemed to still be pushing. I always keep an eye on the other players and I feared they might be ready or very close. Sometimes you want your safest tiles for the very end when everyone will be tenpai.

Tha_Reaper
Fresh Reacher
Fresh Reacher
Posts: 19
Joined: Mon May 18, 2009 10:18 am

Re:game recording for analysis. Good or bad idea?

Post by Tha_Reaper » Wed May 20, 2009 11:56 am

nice video hotelFSR. the only thing that bothered me was the youtube logo that was right in front of the drown tile :P

definitely learned something about reads. and strategic waits.

But... i need to ask something. at a certain moment, you ditched 2 1-pin tiles, and in the beginning of a game, you ditch almost all 9s and 1s. Why? isnt a hand with terminals worth more fu then a hand without them?

Yttrasil
Fresh Reacher
Fresh Reacher
Posts: 49
Joined: Sat Jan 17, 2009 11:11 am

Re:game recording for analysis. Good or bad idea?

Post by Yttrasil » Wed May 20, 2009 12:44 pm

Cool with some more of these videos!

I did quite a few about a year ago and I hope Senechal that I wasn\'t the one you so disliked since I though they were at least decent since I\'ve done quite a few for poker too. =) But what do I know, sadly enough they are gone since I uploaded them onto Stage6 and megaupload and all are gone... might do a new soon if anyone wants to watch and mine weren\'t the hated ones ;)

HotelFSR
Expert Reacher
Expert Reacher
Posts: 328
Joined: Sat Jan 17, 2009 11:39 am
Location: Federated States of Micronesia

Re:game recording for analysis. Good or bad idea?

Post by HotelFSR » Wed May 20, 2009 1:20 pm

definitely learned something about reads. and strategic waits.

But... i need to ask something. at a certain moment, you ditched 2 1-pin tiles, and in the beginning of a game, you ditch almost all 9s and 1s. Why? isnt a hand with terminals worth more fu then a hand without them?


Ironically, you usually want to be shooting for the least possible fu, not the most. Constant two sided waits and all sequences. Fu is there for squeezing a tiny bit more value out of a more rigid hand.

This is an absolutely key point.

You always want to get rid of those honors and terminals aggressively. Fu have an almost negligible impact on the game in terms of planning. Don\'t forget mangan hands and higher don\'t even count Fu! You always shoot for mangan, anything less you want to fold unless you can block a big hand safely with a quick win. 3-Han hands are the occasional exception. Dealers of course always should try to win.

Having terminals really bites you in the ass. They can be used in the least number of combinations of any number tile, so they will weigh you down big time. More than that, there are no good hand patterns that use them. They are only useful as safe tiles unless they directly contribute to your shan-ten (steps to ready) count.

Don\'t forget that having no terminals or honors gives you 1 han! That\'s massive. Forget about fu. It\'s so much more flexible and stackable, just look at player hand stats. Only use honor sets when you have a big lead you want to keep by ending the game faster, otherwise you want to avoid opening your hand because you lose value and safe tiles.

The default pattern you should always shoot for is all sequences with no terminals or honors. It makes use of as many two sided waits as possible and can easily get you into mangan territory when combined with reach, doras or red fives.

Terminals are extra bad when red fives are involved, since using them in a set is mutually exclusive. Unless, of course, you want safe tiles.

Bear in mind that pinfu (all sequences) is about 20 times more common than outside hands (terminal or honor in every set)!!! This is enormous. Outside hands are also difficult to stack with anything else. Go for the easy, stackable hands at all times, and focus on the dora.

Go for the hand value and never discard tiles around the dora unless you have a great hand already and have no possible use for them.


To sum up: honors and terminals are the devil. Only keep honors if you have several matching ones or you have a weak hand and want to withhold them from other players wanting a quick win. Do keep your double-wind tiles if possible, but terminals just plain suck unless you have tons of them already in mostly complete sets.

Tha_Reaper
Fresh Reacher
Fresh Reacher
Posts: 19
Joined: Mon May 18, 2009 10:18 am

Re:game recording for analysis. Good or bad idea?

Post by Tha_Reaper » Wed May 20, 2009 1:52 pm

thats pretty clear :) thanks.

HotelFSR
Expert Reacher
Expert Reacher
Posts: 328
Joined: Sat Jan 17, 2009 11:39 am
Location: Federated States of Micronesia

Re:game recording for analysis. Good or bad idea?

Post by HotelFSR » Wed May 20, 2009 2:05 pm

I went off there, huh :)


The reason I went into such detail is because following that reasoning is probably the single biggest thing anyone can learn to start moving out of the beginner category.

Tha_Reaper
Fresh Reacher
Fresh Reacher
Posts: 19
Joined: Mon May 18, 2009 10:18 am

Re:game recording for analysis. Good or bad idea?

Post by Tha_Reaper » Wed May 20, 2009 4:36 pm

ok, my video is on its way. Its uploading now.
I noticed that the PCM codec for sound is awefull... i ended with a 10 MB video stream and a 136 MB audio stream (making the total video 146 MB ), which is stupid of course.

so i reencoded the audio to mp3, and now i have a nice and small 17 MB video with 21 minutes of gameplay.
I tried to comment on the choices that i made, but hell thats hard :P i think ive been swearing a couple of times too while i thought i had turned off my microphone.


EDIT:
and its done uploading:
http://www.sendspace.com/file/hn7k4z

some comments:
1: at the end i say: his is my second time winning... ever.
I meant there: at tenhou
2: this game was pretty non-standard. it seemed like everyone tried to rush their hands, giving me quite a hard time to try to get a read.
3: at the end, the video will lag around 1 second behind the audio... but no real problems.
4: i play games without red fives!

Bunta
Fresh Reacher
Fresh Reacher
Posts: 37
Joined: Sat Jan 17, 2009 10:16 am

Re:game recording for analysis. Good or bad idea?

Post by Bunta » Thu May 21, 2009 11:08 am

reaper i watched your game and now i say something about it.

First game was okay, although an open kan with a closed hand is not such a good idea, cuz you still have 3 east which is 2 han/yaku so it doesnt help you getting closer to tenpai. keeping your hand concealed is very important cuz you have much more flexibility and can get more points.

2nd hand as dealer. Never ever through east as first cuz it gives you 2 han and it was even the dora. if you kept it west wouldnt win on your discard. East in east round and as dora, safe this tile unless you can reach, then you can throw it!
At 3.54 min in the same round, throw the 1+2 dots and keep the 5 man cuz it gives you more options to win this hand.At 4.19 pon the haku for a quick game to stay dealer.

In the next round, the south pon chun and hatsu so he is going for a fast hand. Maybe he is alread tenpai, so change to defensiv and discard your man tiles, cuz he sure wants a half flush.

Next one: at 10.30 i would have thrown the 7 man cuz it gives two chances of 3 colored straight with 123 and 456. the 445 shape is quiet good so keep it longer and dont brake it so early.
But you kept the 7 and got lucky to get six and eight. But then dont reach, at least wait to get an ryanmen wait(open ender) to get more points for pinfu. okay you won, but it could have been much more pionts!

next, at 12.05 keep the dragons and east for maybe half flush and discard 1man at first. In this game bill is east so he is going for a fast hand after calling the chun. at 14.37 you should have switch to defense and not discard the 9 sou but the 7 cuz its safe for the dealer and the south cuz they look much like a ready hand with almost 2 open sequences. So it was right to break up your hand but you threw the wrong tiles. 7 sou, 5 sou, just discard tiles that have been recently discarded cuz the others are temporaly furiten on those(the tiles of person sitting in front of you are 100% safe against ron).

next hand. Your always doing the same mistake. If you wanna get rid of the honours, then first throw the useless ones for you. So winds that are not round and position wind. So first north, then if you want your own which is south. In the beginning of a hand always keep an eye on which tiles are dora and what wind you are. At 18 min why you keep the 9 dots instead of the 1 man? You have the possibility of full straight in man and maybe 3 colored run 123, so keep it!

So that was it for now. This is only my opinion, so maybe some guys think diffrent about the hands, but i hope it will at least help you a bit.

HotelFSR
Expert Reacher
Expert Reacher
Posts: 328
Joined: Sat Jan 17, 2009 11:39 am
Location: Federated States of Micronesia

Re:game recording for analysis. Good or bad idea?

Post by HotelFSR » Thu May 21, 2009 3:21 pm

I watched it and agree with all of Bunta\'s points.

Three of the biggest mistakes:

Not stealing the haku when you were the dealer and were under threat from another player.

Not going for the half flush when you were dealer/kanning the east tiles. Keep it closed unless you have to open!

The lucky hand you reached on. It may have won you the game, but follow Bunta\'s advice on that one.



I\'ve also made a new video I\'ll be posting in a new thread!

Tha_Reaper
Fresh Reacher
Fresh Reacher
Posts: 19
Joined: Mon May 18, 2009 10:18 am

Re:game recording for analysis. Good or bad idea?

Post by Tha_Reaper » Thu May 21, 2009 3:22 pm

good pieces of advice.

Its true that i still have to learn to look at the wind tiles. as it is now, i only keep those when i have a pair already. With that pair i hope for a triplet, or a open pon so i can end a game fast if thats needed. I havent looked at the player winds yet. I\'m going to try that in the future though.

1: the open kan... I did that because i thought it could give me more points with a second dora, plus i wanted to end it quickly there. Probably wasnt the best choise at that moment.
I have the tendancy to declare all open kan that im able to. i have to try to keep that urge in check :P

2. at the point that i trwe the east tile i didnt even realise it was the dora. and with the round/player winds... see comment above.

on the closed wait around 11:14. I dont know why i decided to wait on that one tile, but for some reason, when i do that it has never failed so far. i think you can call it beginners luck... If i feel uncertain about an open wait, i dont do it. I just looked at the board, and a lot of tiles around the 5 pin were thrown already, but no 5 pins. so i hoped noone had a pon of 5 pins, and that the people didnt save tiles around the 5, making it useless to keep, and thus trow it away.

at 12.05: should i have gone for a Hon Itsu in that situation? Shoudnt i have had at least one pair of the dragons/wind?

last situation: i totally overlooked the straight. good point. I\'ll keep my eyes open.

HotelFSR
Expert Reacher
Expert Reacher
Posts: 328
Joined: Sat Jan 17, 2009 11:39 am
Location: Federated States of Micronesia

Re:game recording for analysis. Good or bad idea?

Post by HotelFSR » Thu May 21, 2009 3:55 pm

Making an early open kan won\'t help you win any faster, and the hope of extra dora is pointless because it will also be given to your opponents! Kans will only help increase/reduce turns when they are called very late in the game.


You will almost always make a higher score with a closed hand. Opening your hand also gives the signal that you are dangerous, and is not worth the tiny bonus of a kan. Keep those easts as safe tiles or to hide the value of your winning hand so that people don\'t fold to you.


About the reach, waiting on one tile is not necessarily bad. Sometimes it can even be a good thing if you wait on a tile that looks safe based on double sided wait logic. However in this particular case it was a bad idea because 5 is an unlikely wait and because your hand was worth almost nothing. You might have dealt into a much bigger hand, with nothing to gain in return.

By the way, when you see no fives on the board that means you SHOULDN\'T wait for them because they are probably needed or being used. Wait for tiles that you can see on the board, because it means it\'s much more likely your opponents don\'t need them and will throw them to you when pushing.

You cannot assume that if people have thrown tiles around the five that they don\'t need it. In fact, it\'s usually the opposite. That\'s where the suji reading concepts come from, which I\'ll cover more closely in a video at some point. e.g. in your last hand you threw a green 5 but were waiting on a 6!

Those players seemed to be low level so they weren\'t careful with their tiles, giving you the 5 and the 6 in the last hand without attempting to fold. It\'s important to develop good habits that will serve you well against players of all levels.

Yttrasil
Fresh Reacher
Fresh Reacher
Posts: 49
Joined: Sat Jan 17, 2009 11:11 am

Re:game recording for analysis. Good or bad idea?

Post by Yttrasil » Thu May 21, 2009 5:11 pm

So just watched the 2nd video closer the first one I didn\'t really pay attention to HotelFSR, anyhow you need to work on your game around 3-5 min into it many many of the throws you did there were just bad so I recommend everyone not to play like that. Though I have to say that it\'s a great video for beginners to begin thinking about their own play and analyze hands also very fun to listen too. It\'s like some pokervideos you watch and even though it sounds great there are quite a few mistakes in the thoughts behind the actions and therefore some bad play from time to time. Looking forward to more though for sure ;)

Post Reply